How Big Is the Quantum Defence Sector?

The biggest number in quantum defence is $151 billion. It is the ceiling on SHIELD, a Missile Defense Agency contract held by more than 2,400 firms, among them IonQ, Infleqtion, Vector Atomic and QuSecure. A ceiling is the most the government may spend across every holder, over a contract that can run to 2035. So far each of the four has a single order of $500, the same token sum almost every holder received for signing on.

That gap is why the size of the quantum defence sector gets quoted at anything from a few million dollars to a few billion. Most of those figures are not wrong. They answer different questions, and this guide keeps them apart.

The figures that can be checked are modest. Defence took between a quarter and two-fifths of all federal quantum research money each year from FY2019 to FY2025. Its disclosed contracts came to $114 million to $141 million a year from 2023 to 2025, and only paid market reports reach into the billions.

This year looks bigger, but for a narrow reason. Disclosed contracts for 2026 had reached $191 million or $223 million by 11 October, depending on how one PsiQuantum agreement is read. Two announced DARPA deals, with PsiQuantum and Diraq, explain all of the rise, and without them 2026 so far comes to about $47 million, lower than any earlier year.

Four ways to measure the sector, and why they cannot be added

There are four ways to put a number on defence quantum money. Each answers a different question. A budget line is what a government plans to spend, often in its own laboratories, while a market sizing is an analyst’s estimate of what all buyers spend.

A contract award is money promised to a named firm or university for a named job. A revenue share is what a listed company tells its investors about who pays its bills.

The commonest mistake is to add them up. A Pentagon research budget pays for contracts, and those contracts turn up again as company revenue, so the sum counts one dollar three times. Each measure also hides a trap of its own, because a request is not a budget, a budget is not money paid out, and a ceiling is not an award.

The Pentagon’s quantum research budget was about $320 million in 2023

The best single figure comes from the federal National Quantum Initiative report for FY2025, published in December 2024. One of its charts shows the quantum research budgets of five agencies, but the report prints no agency numbers, so ours are read off the bars. They are good to about $3 million either way.

The defence bar shows about $320 million of actual spending in FY2023. It rises to about $400 million estimated for FY2024, then falls to about $280 million requested for FY2025.

Against the whole defence research budget, that is a sliver. The FY2025 request was 0.2 per cent of the Pentagon’s $143 billion request for research, development, test and evaluation, set out in the FY2025 act’s funding tables. The report does not say whether classified work is included.

Within federal quantum research, though, defence is a big customer. All federal agencies together spent $1,036 million on quantum research in FY2023. Defence took between about 27 and 38 per cent in FY2019 to FY2023, about 39 per cent in the FY2024 estimate and about 28 per cent in the FY2025 request. The report calls quantum one of the Pentagon’s 14 critical technology areas, with steady defence funding for more than 30 years.

Only one budget line in the last two defence authorisation acts has quantum in its name. Line 0603330D8Z, “Quantum Application”, was authorised at $69 million for FY2025 in the FY2025 act and $60 million for FY2026 in the FY2026 act. But authorised is not spent. FY2025 ran on a full-year stopgap budget, so the amount actually appropriated for either year is unconfirmed.

Congress added a little more through lines with other names. For FY2025 it authorised an extra $10 million for DARPA’s US2QC quantum computing programme and $7.5 million for Air Force quantum materials. For FY2026 it authorised $45 million for four quantum projects tucked inside three broader research lines, such as the Air Force’s Defense Research Sciences.

The biggest Pentagon numbers are caps, not spending. The FY2025 act wrote DARPA’s Quantum Benchmarking Initiative into law without a dollar figure. Its programme page says each firm can win up to $1 million in its first stage, $15 million in its second and $300 million in its third. No total is published.

Outside the US, defence quantum money shows up only in pieces

The US is the only country where we found a dated, yearly figure for defence quantum spending. Our search elsewhere was limited. We checked Britain, France, Canada, Japan for FY2025, Australia, the European Union, NATO and China, but not Japan’s FY2026 budget, South Korea, Israel, India, Germany or the AUKUS pact.

Britain’s Ministry of Defence says it has put more than £28 million into quantum research and development through Dstl, its science laboratory (gov.uk, 2 January 2025). It does not say over what period. Eight Dstl and ministry quantum contracts listed on Contracts Finder, awarded between May 2022 and November 2023, came to about £948,000 in total. The newer Find a Tender service shows none at all from 24 February 2025 to 1 October 2026, though an empty notice board does not prove nothing was spent.

The European Defence Fund backs ADEQUADE, a 36-month, €27 million defence research project on quantum sensing that its factsheet says Thales leads. The European Defence Agency issued a €1.6 million tender in 2025 for a defence action plan and foundational research on quantum technologies.

Canada promised more than $20.3 million over two years, in Canadian dollars, for a quantum defence hub at the University of Calgary. Japan’s defence ministry set aside ¥2.0 billion for FY2025 to test new ways to keep messages safe, from quantum key distribution to other new kinds of encryption.

Australia’s defence accelerator, ASCA, signed 21 contracts worth a combined A$60 million in late 2024, trade press reported. It did not say how much went to quantum. The NATO Innovation Fund led a €5 million seed round in Aquark Technologies, which makes cold-atom sensors, but did not say how much of it the fund paid.

France’s PROQCIMA programme, which buys quantum computers for defence, has a ceiling rather than a budget. Its framework agreements allow a maximum of €500 million over at least 10 years, with €60 million for phase 1 and up to €150 million for phase 2. Pasqal left the programme after the first phase, it said on 28 September 2026.

NATO approved a quantum roadmap on 30 July 2026 with no money attached. We found no credible defence quantum figure for China. None of these numbers can be added to the American one, because they come in different currencies, cover different years and mix caps with signed deals.

Paid market reports disagree by a factor of nearly ten

The big consultancies do not put a dollar figure on defence. McKinsey’s 2025 Quantum Technology Monitor comes closest. It says government, defence included, held 62 to 66 per cent of a quantum communication market worth about $0.9 billion to $1.0 billion in 2023. BCG puts the whole quantum computing market at about $550 million in 2025, more than half of it spent by businesses.

Most headline numbers come from paid reports on “quantum warfare”. They cannot agree. For 2024 they range from DataM’s $196 million and MRFR’s $203 million to Custom Market Insights’ $1.5 billion and MarketsNXT’s $1.9 billion, all for the same named market.

Later years split the same way. Research Nester puts 2026 at $290 million, while The Business Research Company puts 2025 at $1.45 billion.

Even the headlines are loose. DataM’s own press release says “$1,578.1 Billion” in its headline where the text below says million, and the methods behind all of these reports are sold rather than published.

Averaging them would be a mistake. The reports fall into two groups, one near $0.2 billion and one between $1.5 billion and $1.9 billion, and a mean would land in the gap at a figure no analyst ever estimated. The fair summary is a range of $0.2 billion to $1.9 billion for 2024, taken from the free teasers.

Defence contracts rose in 2026, but only on two DARPA deals

Contracts name who gets paid. The money comes in two kinds, and they should not be confused. Obligated money is recorded on USAspending, the federal register of contracts and grants, and it is the firmest figure anyone can check. Announced money is a value a company gives in a press release or stock market filing, and it may never show up in the register.

Our totals are a floor. A keyword search of the whole register finds more sensing and communications money than our list of named recipients does.

Outside computing, the money goes to sensors, clocks, networks and codes. The register shows about $35 million of such awards starting in 2023, $48 million in 2024, $24 million in 2025 and $2.7 million so far in 2026. Announced money missing from the register adds nothing in 2023, then $55 million in 2024, $30 million in 2025 and $32 million in 2026.

In 2026 that announced money is mostly IonQ’s clock contract. Before then it is mostly networking. IonQ reported a $55 million Air Force contract in 2024 for “the scaling, networking, and deployability of quantum systems”, and Rigetti a $5.8 million Air Force networking contract in 2025. Either could count as computing, so our computing guide shows the totals both ways.

Some of the money buys ordinary software, not quantum hardware. We label it rather than drop it. Awards to QuSecure and Terra Quantum for post-quantum cryptography, code built to resist a future quantum computer, come to $9.9 million over the four years. Two Infleqtion awards for machine-learning software add $3.0 million. QuSecure’s $3.9 million in 2026 is announced, and may amend its 2024 Air Force contract rather than be a new award.

The names behind the totals matter. In 2023 Montana State University took $33 million of the $35 million on the register outside computing. It came in two Air Force grants, $27 million for its QCORE test site and $6.2 million for a testbed that links quantum devices.

The only company money on the register that year was QuSecure’s $1.7 million for code that resists quantum attack. Montana State’s QCORE grant is also separate from a 2025 Air Force deal the university put at $32 million, of which about $16 million is on the register so far.

In 2024 it was clocks and networks. Vector Atomic won an $11 million Army contract and a Navy contract with $10 million obligated, or $16 million once all its options are counted. Infleqtion won an $11 million Army contract for optical clocks that fit in a rack.

Qubitekk, now part of IonQ, won $9.1 million from the Air Force for quantum network work. With all options it comes to $21 million, which matches the $21 million project IonQ announced in January 2025, so we count the two as one award.

The largest award outside computing that we found is IonQ’s $55 million of 2024. It comes from an SEC filing and is not on the register. The largest for clocks is IonQ’s $28 million DARPA contract extension of 2026, by IonQ’s own account, and a further $30 million option has not been taken up, so we do not count it.

For navigation sensors the largest is Q-CTRL’s. DARPA chose it for two navigation sensor contracts “valued at A$38M (US$24.4M)”, announced on 27 August 2025. Those are not on the register either. Many awards carry no price at all, such as Safran’s place in the same DARPA programme and the prototype quantum navigation system Lockheed Martin won from the Defense Innovation Unit.

Quantum computing took $80 million in 2023, $38 million in 2024 and $67 million in 2025, counting both kinds of money. Then 2026 jumped. By 11 October it had reached $157 million or $188 million, roughly double to triple the 2023 to 2025 average, and all of the rise is announced money from two DARPA deals. They are PsiQuantum’s $125 million agreement and Diraq’s initial US$51 million funding tranche.

Obligated computing money went the other way. It fell from about $55 million in 2023 to $35 million in 2025 and $12 million so far in 2026, though the register runs behind. Our guide to defence money for quantum computing has the detail.

The Missile Defense Agency’s $151 billion SHIELD ceiling is left out of every figure here. SHIELD is an IDIQ, short for indefinite-delivery, indefinite-quantity. Such a contract fixes the terms of future work without fixing how much will be bought, so the agency signs up suppliers first and buys from them one order at a time.

Federal rules require every contract of this kind to guarantee a minimum order. On SHIELD that minimum looks like a formality. Of the 2,401 initial orders on the register, 2,395 are for exactly $500, and the whole batch comes to about $1.2 million against a ceiling of $151 billion.

The four quantum holders we checked, IonQ, Infleqtion, Vector Atomic and QuSecure, each have one of those $500 orders. The sum predicts nothing. Real money comes through task orders, individual jobs that the agency and other Pentagon bodies put out to competition among holders.

Until then no other money is committed. The Pentagon’s award notice says no funds will be obligated on the base award, only at the order level, and none of the four quantum firms had won a task order by late October.

Government pays most of the bills at some quantum firms, but none splits out defence

Listed quantum firms have to tell investors who their big customers are. Rigetti’s annual report says US government bodies supplied 48 per cent of its 2025 revenue, and governments of every kind 90 per cent, up from 89 per cent the year before. On revenue of about $7.1 million, the US government share comes to about $3.4 million.

Infleqtion’s annual report puts US government sales at about 60 per cent of FY2025 revenue and British government work at about 12 per cent. Quantinuum leans on Washington far less. Its listing prospectus says the US government provided 16 per cent of FY2025 revenue and 24 per cent in the first quarter of 2026. Japan’s RIKEN research institute provided 60 per cent of FY2025 revenue.

IonQ gives no government share. It says only that three unnamed customers made up 53 per cent of its 2025 revenue. D-Wave gives none either, but its annual report shows only about $2.7 million of its $25 million revenue in 2025 came from the United States, against $17 million from Germany. That caps what it could have earned from American defence customers, including Davidson Technologies, which has hosted a D-Wave machine in Alabama since November 2025.

Government is not the same as defence. No filer separates the Pentagon from civil agencies such as the Department of Energy and NASA, so these shares show how much the firms lean on public money. They do not show how much of it is military.

What the armed forces actually buy is clocks

The contracts show what the armed forces are paying for, and it is mostly clocks and navigation sensors. These keep time and position when satellite signals are jammed, which is why the Army, the Navy and the Department of the Air Force have all bought them since 2024.

Clocks are moving into production. Vector Atomic’s Army award is a third-phase small-business contract to scale up production of timing systems. DARPA has since hired IonQ, which bought Vector Atomic, to set up a pilot production line for tactical optical clocks. IonQ says its $28 million modification “will cover manufacturing development and 25 Evergreen-05 clocks”, and a further $30 million option, “not yet exercised”, would cover 100 more.

Quantum computers have been bought too, but only as research hardware. An Air Force laboratory bought two IonQ systems in 2023 for networking research, and Britain’s defence ministry bought one small ORCA machine in 2022 at a price it never gave. No armed force has publicly ordered a quantum computer for use in operations.

Which numbers to quote, and which to ignore

Investors sizing a firm’s defence business should stop quoting the $1.5 billion to $1.9 billion market reports. Nobody outside the publishers can check them. Nor is the $151 billion SHIELD ceiling an order book, since the quantum firms on it hold only a token $500 first order each. A firm’s own filing gives its government share, and the federal register gives the awards it can name.

Anyone quoting the size of the sector should give the measure, the year, the country and the status, since money can be requested, enacted, obligated, announced or merely capped. Each gives a different number. The best-supported single figure is the Pentagon’s roughly $320 million of quantum research spending in FY2023.

That figure, read off a bar chart, is why we say hundreds of millions rather than billions. The verdict would be wrong if a later budget crosscut or a Pentagon budget book showed defence quantum research several times the FY2024 estimate.